PSU measuring
#1
I have a system currently running just 8gb of ram, and a pentium 3.3 dual core CPU.  In that system I also have 1 GTX 1070 and 1 GTX 1080, both founders edition.  My current PSU is a Corsaire 1000 watt, and many have told me on these forums that if I wanted to put a 3rd GPU in there I would most certainly need a bigger PSU.  So I wanted to measure my current pull on the PSU and bought one of these:  

http://a.co/2iIYEVe

Now, in idle state that meter reads about 100watts, and when I start pushing my GPUs using hashcat, the meter jumps up to about 300w.  My question is this.  Does that accurately show my current pull on the psu or merely the pull from my wall.  Even if another GPU pulled another 200w wouldn't that only put me at half of the PSU's capacity, or am I thinking about this wrong..  Educate me please.

I guess the question is:
Is there a way to read the pull on the PSU and not the wall outlet or are they the same?
Reply
#2
The pull from the wall should be just a little bit higher than the pull from the PSU due to losses from non-perfect efficiency. Under worst case scenario, your 1080 will draw a rough 300W and your 1070 will draw maybe a peak of 200W. Add another 200W for the rest of the board just to be safe and you still, under the calculated worst case draw, have ~300W to use. You can fit a whole second 1080 onto that machine with the current power supply and be relatively safe.
Reply
#3
Make sure you are testing power consumption while running something like single hash MD4 brute force.
Reply
#4
(05-16-2017, 06:39 PM)epixoip Wrote: Make sure you are testing power consumption while running something like single hash MD4 brute force.

Very good point I will run that when I get home.. I was testing -o 2500 wpa
Reply
#5
(05-16-2017, 06:42 PM)elidell Wrote:
(05-16-2017, 06:39 PM)epixoip Wrote: Make sure you are testing power consumption while running something like single hash MD4 brute force.

Very good point I will run that when I get home.. I was testing -o 2500 wpa

Try with workload profiles 3 and 4 as well if you aren't already:

.. -w 3
.. -w 4
Reply
#6
(05-23-2017, 03:36 PM)kjs Wrote:
(05-16-2017, 06:42 PM)elidell Wrote:
(05-16-2017, 06:39 PM)epixoip Wrote: Make sure you are testing power consumption while running something like single hash MD4 brute force.

Very good point I will run that when I get home.. I was testing -o 2500 wpa

Try with workload profiles 3 and 4 as well if you aren't already:

.. -w 3
.. -w 4

I have , and quite frankly I don't notice a difference.. not in power, not in hash rat, not in time.. nothing..  I don't think I understand what -w does. In fact the only thing that seems to increase my hash rate is a larger word list..smaller word list drop my hash rate down almost in half.. I assume this has something to do with hashcat optimizing the work load.

Please help me understand the -w option
Reply
#7
- [ Workload Profiles ] -

# | Performance | Runtime | Power Consumption | Desktop Impact
===+=============+=========+===================+=================
1 | Low | 2 ms | Low | Minimal
2 | Default | 12 ms | Economic | Noticeable
3 | High | 96 ms | High | Unresponsive
4 | Nightmare | 480 ms | Insane | Headless

It definitely increases performance when I specified profiles 3 and 4 (in the two systems I've tested so far with 8 x 1080 Ti's and 11 x RX 580's).
Reply